Document Type : Research Article
Department of Civil Engineering, Najafabad Branch, Islamic Azad University, Najafabad, Iran
The use of self-compacting concrete types is increasing day by day, and understanding its rheological behavior is a high priority in the use of this type of concrete. In this study, the rheological properties of fiber-reinforced ultra-high performance self-compacting concrete (UHPSCC) were predicted by ANN-GA (Genetic Algorithm hybrid with Artificial Neural Network) and RBF-NN (Radial Function Neural Network), and its output was compared with laboratory results. The purpose of this study is to evaluate the making of UHPSCC with durable and to achieve an artificial neural network that can more accurately predict the rheology properties of this type of concrete. Rheology properties of self-compacting concrete include tests related to fresh concrete, including; Slump flow (D), slump flow (T50), V-funnel flow, and L-box test were performed. Experimental results indicate high compressive and tensile strength and rheological properties placement within the acceptable EFNARC range. Estimation and prediction of the two neural networks studied from the rheological properties of this type of concrete show the acceptable accuracy of prediction of both neural networks. Between these two artificial neural networks, the prediction accuracy of ANN-GA is higher.